Stacked sparse autoencoder in hyperspectral data classification using spectral-spatial,higher order statistics and multifractal spectrum features

This paper proposes a novel classification paradigm for hyperspectral image (HSI) using feature-level fusion and deep learning-based methodologies. Operation is carried out in three main steps. First, during a pre-processing stage, wave atoms are introduced into bilateral filter to smooth HSI, and this strategy can effectively attenuate noise and restore texture information. Meanwhile, high quality spectral-spatial features can be extracted from HSI by taking geometric closeness and photometric similarity among pixels into consideration simultaneously. Second, higher order statistics techniques are firstly introduced into hyperspectral data classification to characterize the phase correlations of spectral curves. Third, multifractal spectrum features are extracted to characterize the singularities and self-similarities of spectra shapes. To this end, a feature-level fusion is applied to the extracted spectral-spatial features along with higher order statistics and multifractal spectrum features. Finally, stacked sparse autoencoder is utilized to learn more abstract and invariant high-level features from the multiple feature sets, and then random forest classifier is employed to perform supervised fine-tuning and classification. Experimental results on two real hyperspectral data sets demonstrate that the proposed method outperforms some traditional alternatives.     

一种双正则项全变差高光谱图像去噪算法

受传感器特性影响,高光谱图像中的噪声在各个维度都有体现。噪声的存在降低了高光谱图像中信息的有效性,在进行地物分类前必须采用适当的算法对噪声予以去除。文章针对高光谱图像的噪声特性,提出了一种基于全变差的高光谱图像去噪算法。该算法将经典二维图像全变差去噪模型推广至三维形式,提出了采用双正则项及相应的调整参数的目标函数,在三维空间中完成新目标函数的离散化,并采用基于优化-最小化算法的迭代方法对目标函数进行优化与求解。对星载Hyperion成像光谱仪数据的实验表明,适当的设置调整参数,该方法可很好地提高高光谱图像的各波段信噪比、平滑光谱曲线并保留细节特征,其去噪效果优于经典的MNF去噪算法及Savitzky-Golay滤波方法。     

Wavelet SVM in Reproducing Kernel Hilbert Space for hyperspectral remote sensing image classification

Combining Support Vector Machine (SVM) with wavelet analysis, we constructed wavelet SVM (WSVM) classifier based on wavelet kernel functions in Reproducing Kernel Hilbert Space (RKHS). In conventional kernel theory, SVM is faced with the bottleneck of kernel parameter selection which further results in time-consuming and low classification accuracy. The wavelet kernel in RKHS is a kind of multidimensional wavelet function that can approximate arbitrary nonlinear functions. Implications on semiparametric estimation are proposed in this paper. Airborne Operational Modular Imaging Spectrometer II (OMIS II) hyperspectral remote sensing image with 64 bands and Reflective Optics System Imaging Spectrometer (ROSIS) data with 115 bands were used to experiment the performance and accuracy of the proposed WSVM classifier. The experimental results indicate that the WSVM classifier can obtain the highest accuracy when using the Coiflet Kernel function in wavelet transform. In contrast with some traditional classifiers, including Spectral Angle Mapping (SAM) and Minimum Distance Classification (MDC), and SVM classifier using Radial Basis Function kernel, the proposed wavelet SVM classifier using the wavelet kernel function in Reproducing Kernel Hilbert Space is capable of improving classification accuracy obviously.     

Learnable three-dimensional Gabor convolutional network with global affinity attention for hyperspectral image classification

Benefiting from the development of hyperspectral imaging technology, hyperspectral image (HSI) classification has become a valuable direction in remote sensing image processing. Recently, researchers have found a connection between convolutional neural networks (CNNs) and Gabor filters. Therefore, some Gabor-based CNN methods have been proposed for HSI classification. However, most Gabor-based CNN methods still manually generate Gabor filters whose parameters are empirically set and remain unchanged during the CNN learning process. Moreover, these methods require patch cubes as network inputs. Such patch cubes may contain interference pixels, which will negatively affect the classification results. To address these problems, in this paper, we propose a learnable three-dimensional (3D) Gabor convolutional network with global affinity attention for HSI classification. More precisely, the learnable 3D Gabor convolution kernel is constructed by the 3D Gabor filter, which can be learned and updated during the training process. Furthermore, spatial and spectral global affinity attention modules are introduced to capture more discriminative features between spatial locations and spectral bands in the patch cube, thus alleviating the interfering pixels problem. Experimental results on three well-known HSI datasets (including two natural crop scenarios and one urban scenario) have demonstrated that the proposed network can achieve powerful classification performance and outperforms widely used machine-learning-based and deep-learning-based methods.     

Brain tissue classification based on DTI using an improved Fuzzy C-means algorithm with spatial constraints

We present an effective method for brain tissue classification based on diffusion tensor imaging (DTI) data. The method accounts for two main DTI segmentation obstacles: random noise and magnetic field inhomogeneities. In the proposed method, DTI parametric maps were used to resolve intensity inhomogeneities of brain tissue segmentation because they could provide complementary information for tissues and define accurate tissue maps. An improved fuzzy c-means with spatial constraints proposal was used to enhance the noise and artifact robustness of DTI segmentation. Fuzzy c-means clustering with spatial constraints (FCM_S) could effectively segment images corrupted by noise, outliers, and other imaging artifacts. Its effectiveness contributes not only to the introduction of fuzziness for belongingness of each pixel but also to the exploitation of spatial contextual information. We proposed an improved FCM_S applied on DTI parametric maps, which explores the mean and covariance of the feature spatial information for automated segmentation of DTI. The experiments on synthetic images and real-world datasets showed that our proposed algorithms, especially with new spatial constraints, were more effective.     

Motion detection in moving background using a novel algorithm based on image features guiding self-adaptive Sequential Similarity Detection Algorithm

A novel efficient algorithm for motion detection in dynamic background was proposed. In image registration step, a feature-based and self-adaptive Sequential Similarity Detection Algorithm (SSDA) algorithm was proposed, which searches for matching position under constraints induced by image features with variational threshold. Then perform change detection by calculating and classifying the Mean Absolute Difference (MAD) around detected features in the middle frames of three consecutive images. Moving objects position was determined according to the rule that the feature from moving regions shows a lager MAD. Experiments on data sets of four typical scenes show that the improved registration algorithm is accurate and costs less than 0.4 s in computation, much faster compared with other four methods, and the proposed Dual Maximum Mean Absolute Difference Algorithm (DMMADA) can obtain a robust set of moving object features. Our algorithm can be used for fast detection of moving targets in dynamic background as well as change detection.